JP4107733B2 - B-mode processing apparatus for ultrasonic imaging system and post-detection image processing method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates generally to ultrasound imaging of human tissue and blood. Specifically, the present invention relates to a method for improving the quality of B-mode ultrasound images.
[0002]
BACKGROUND OF THE INVENTION
A typical ultrasound imaging system has an array of ultrasound transducers that are used to transmit an ultrasound beam and receive a reflected beam from an object to be examined (subject). For ultrasound imaging, the transducer array typically has a number of transducers arranged in a row and driven by separate voltages. By controlling the individual transducers by selecting the time delay (or phase) and amplitude of the applied voltage, the ultrasound generated by these transducers is combined along the preferred vector (beam) direction. A net ultrasonic wave can be formed that travels and converges to a selected point along the beam. Multiple firings can be used to obtain data representing the same anatomical information. Changing the beam forming parameters of each firing, changing the maximum focus, or transmitting beams one after another along the same scan line, eg, shifting the focus of each beam relative to the focus of the previous beam Thus, the contents of the received data can be changed for each firing. By changing the time delay and amplitude of the applied voltage, the beam can move with its focal point in the plane for scanning the object.
[0003]
The same principle applies when the transducer is used to receive reflected sound waves (receive mode). The voltage generated by the receiving transducer is added so that the net signal represents the ultrasound reflected from one focal point in the subject. As with the transmit mode, this focused reception of ultrasound energy is achieved by providing separate time delays (and / or phase shifts) and gains for the signals from each receive transducer. .
[0004]
FIG. 1 shows a conventional ultrasound imaging system having four subsystems: a beam former 2, a B-mode processor 4, a scan converter / display controller 6, and a kernel 8. System control is centralized in the kernel 8, which accepts operator input via the operator interface 10 and controls the various subsystems. The main controller 12 performs system level control functions. This accepts operator input and system state changes via the operator interface 10 and makes appropriate system changes either directly or via a scan controller. System control bus 14 constitutes the interface from the main controller to the subsystem. A scan control sequencer 16 provides real time (acoustic vector rate) control inputs to the beamformer 2, the system timing generator 24, the B-mode processor 4 and the scan converter 6. The scan control sequencer 16 is programmed by the host to have the option of vector sequence and synchronization for acoustic frame acquisition. The scan converter sends the vector parameters defined by the host to the subsystem via the scan control bus 18.
[0005]
The main data path begins with an analog RF input from the transducer 20 to the beamformer 2. The beam former 2 converts the analog signal into a series of digital samples and outputs two summed digital receive beams. These outputs are shown as complex I and Q data in FIG. 1, but may generally be radio frequency or intermediate frequency data. The I and Q data are input to the B-mode processing device 4 and output to the scan converter / display controller 6 as processed vector (beam) data. The scan converter accepts the processed vector data and outputs a video display signal to the color monitor 22 for displaying the image.
[0006]
Referring to FIG. 2, a conventional ultrasound imaging system includes a transducer array 20 having a plurality of separately driven transducer elements 26, each transducer element 26 being generated at the transmitter of beam forming apparatus 2. A burst of ultrasonic energy is generated when energized by a pulse waveform. The ultrasonic energy reflected from the subject and returning to the transducer array 20 is converted into an electric signal by each receiving transducer element 26 and is transmitted to the beam forming apparatus 2 via a pair of transmission / reception switching (T / R) switches 28. Are separately applied to the receiver. The T / R switch 28 is typically configured with a diode to protect the receiver electronics from the high voltage generated by the electronics of the transmitter. A diode is activated by the transmission signal to block or limit the signal to the receiver.
[0007]
The transmitter and receiver of the beam forming device 2 are operated under the control of a beam forming device controller (not shown) in response to commands from the operator. The elements of transducer array 20 are driven so that the generated ultrasonic energy is directed or steered in the form of a beam. To accomplish this, each pulse generator 30 is given a respective time delay. Each pulse generator 30 is connected to a respective transducer element via a T / R switch. The time delay for transmission focusing is preferably read from the lookup table 32. By appropriately adjusting the time delay for transmission focusing as usual, the ultrasonic beam is directed away from the axis Y by an angle θ and focused at a certain distance (range) R. Sector scanning is performed by progressively changing the time delay for the transmission focus of successive excitations. Accordingly, the angle θ is changed incrementally and the transmitted beams are steered in successive directions.
[0008]
An echo signal is created by each burst of ultrasonic energy reflected from objects located at successive distances along the ultrasonic beam. The echo signal is sensed separately by each transducer element 26 and represents the amount of reflection that a sample of the echo signal amplitude at a particular point in time occurs at a particular distance. However, due to the difference in propagation path between the reflection point P and each transducer element 26, these echo signals are not detected simultaneously and their amplitudes are not equal. The receiving unit of the beam forming apparatus 2 gives an appropriate time delay to each received signal, adds them, and is reflected from the point P located at the distance R along the ultrasonic beam directed in the direction of the angle θ. A single echo signal that accurately represents the total ultrasonic energy. To accomplish this, multiple receive channels 34 are given time delays for their respective receive focus. The time delay for the receive focus is preferably read from the lookup table 38. The receive channel also has circuitry (not shown) for performing apodization and filtering on the received pulses. The time delayed received signals are then summed in receive adder 36.
[0009]
Referring to FIG. 2, the receiving unit of the beam forming apparatus 2 includes a time gain control unit and a receiving beam forming unit. The time gain controller includes a respective amplifier 40 and a time gain control (TGC) circuit 42 for each receive channel 34. Each amplifier 40 is coupled at its input to a respective transducer element 26 and amplifies the received echo signal. The amplification degree of the amplifier 40 is controlled by the TGC circuit 42, and the TGC circuit 42 is set by a manual operation of the potentiometer 44.
[0010]
The receive beamformer includes separate receive channels 34, each receive channel 34 receiving an analog echo signal from a corresponding amplifier 40. Each amplified signal is sent to a pair of quadrature detectors in the respective receive channel, where the mixing reference frequency is 90 ° different. Since this reference frequency is the same as the frequency of the transmission pulse, the output from the low-pass filter in the reception channel is a complex (I and Q) signal having a phase different by 90 °. These signals are output as a series of digitized output values via the I bus 45a and Q bus 45b (or equivalent RF bus). Each of these I and Q baseband signals represents a demodulated sample of the echo signal envelope at a particular distance R. When these samples are summed with the I and Q samples from each of the other receive channels 34 at summing points 36a and 36b, the resulting sum signal is at a distance R along the steered beam (θ). Delayed to represent the magnitude and phase of the echo signal reflected from the located point P.
[0011]
A detector 46 included in the B-mode processor 4 receives beam samples from summing points 45a and 45b. Each beam sample I and Q value is a signal representing the in-phase component and the quadrature component of the magnitude of the reflected wave from the point (R, θ). The detector 46 calculates a quantity (I ² + Q ² ) ^1/2 representing the envelope of the baseband data. If the beam sample is RF data, the signal envelope can be obtained by a standard rectifier followed by a low pass filter. The B-mode function visualizes the time-varying amplitude of the signal envelope in gray scale by some additional processing such as edge enhancement and logarithmic compression (hereinafter referred to as “post-detection image processing”).
[0012]
A scan converter 6 (see FIG. 1) receives display data from the B-mode processor 4 and converts the data into a desired image for display. Specifically, the scan converter 6 converts acoustic image data from a polar coordinate (R-θ) sector format or a Cartesian coordinate linear array into appropriately scaled Cartesian coordinate display pixel data at video rate. The scan-converted acoustic data is then output for display on the display monitor 22, and the display monitor 22 visualizes the time-varying amplitude of the signal envelope in grayscale.
[0013]
Ultrasound imaging has the problem of creating unique imaging artifacts called speckles (spots). A speckle is a spot or spot in an image resulting from the interference pattern of a plurality of received echoes. This spot is mainly caused by nulls in the acoustic interference pattern, but other anomalies in the image, such as irregular electronic noise, can also cause spots. Acoustic zero is accentuated by the logarithmic compression necessary to display the full dynamic range ultrasound image. These acoustic zeros appear as black holes in the image. It is desirable to minimize speckle to improve image quality.
[0014]
Post-detection image processing generally consists of dynamic range (logarithmic) compression, low pass filtering and edge enhancement filtering. They are arranged in different orders on different scanners, but are usually performed sequentially. Traditionally, low pass filters are designed to prevent aliasing prior to data downsampling, but low pass filters also play a role in speckle reduction in wideband imaging systems. For edge enhancement filtering, a high-pass filter that operates on logarithmically compressed data is typically used.
[0015]
The detected image of the actual anatomy usually contains a large reflected signal (from the edge) and a low amplitude speckle (from the soft tissue). Thus, when the low-pass filter and the high-pass filter simply play their role, there is always a tendency for the low-pass filter to blur edges and the high-pass filter to enhance speckle. In the case of a sequential processing configuration, it is very difficult to prevent the low pass filter and the high pass filter from adversely affecting each other. Therefore, the best results cannot be obtained for both speckle smoothing and edge enhancement with opposing effects.
[0016]
4 to 6 show a conventional configuration for sequential post-detection image processing in an ultrasound imaging system. The first configuration shown in FIG. 4 is similar to the post-detection image processing method used in analog systems, and includes a logarithmic compression means 48, a high-pass filter 50 for edge enhancement, and a low-pass filter 52. Have The low-pass filter 52 is typically a fourth-order to sixth-order IIR filter, and the cutoff frequency is set according to a decimation rate before scan conversion. The advantage of this first configuration is that the high pass filter 50 is effective in enhancing the edges of the logarithmically compressed image. If the detected image is logarithmically compressed after high pass filtering, the edge enhancement effect may be reduced by logarithmic compression. The disadvantage of the first configuration is that the low-pass filter after logarithmic compression is used with the intention of preventing aliasing, and is not very effective in reducing speckle.
[0017]
FIG. 5 shows a second configuration. In the implementation of this configuration, a decimation means or rate converter may appear before the high-pass filter 50 for edge enhancement. Therefore, the low-pass filter 52 can perform both speckle smoothing and aliasing prevention. The advantage of the second configuration is that the low-pass filter is optimally placed (before nonlinear compression) for speckle reduction. The disadvantages of the second configuration are that the low-pass filter tends to blur edges, the high-pass filter may emphasize background speckles, and the signal bandwidth is expanded again after logarithmic compression. The position of the low pass filter is not optimal.
[0018]
FIG. 6 shows a third configuration. Similar to the second configuration, decimation or rate conversion appears before high-pass filtering for edge enhancement. The main improvement is the adaptability of the edge enhancement filter, which typically has parallel high-pass filters and all-pass paths. The advantage of the third configuration is that the adaptive edge enhancement filter attempts to distinguish edges from speckle based on the difference in amplitude, and this edge enhancement filter tends to emphasize only edges of large amplitude. That is. The disadvantage of the third arrangement is that the edge is already smeared by the low-pass filter before the processed signal reaches the adaptive edge enhancement filter, and the position of the adaptation mechanism is not optimal, i.e. the edge This is because the difference between the amplitudes of and the speckles has already been significantly reduced by logarithmic compression.
[0019]
According to a fourth configuration (not shown), the RF spectrum is divided into two or more subbands and each subband is detected separately. Adding non-coherent images after detection is an effective method for reducing speckle. However, this type of frequency synthesis has the same statistical performance as the second configuration shown in FIG. Although the addition of the non-coherent image after detection can reduce speckles, the division of the RF spectrum has the same resolution degradation (edge blurring) effect as the low-pass filter in the second configuration.
[0020]
SUMMARY OF THE INVENTION
The present invention is a method and apparatus for adaptive enhancement of B-mode images during post-detection image processing in an ultrasound imaging system. Speckle smoothing and edge enhancement impose incompatible requirements in the design of image processing devices after B-mode detection. Low pass filters that can smooth speckles tend to blur the edges in the detected image. High-pass filters for edge enhancement tend to enhance background speckles or cancel out the effects of smoothing filters. Furthermore, optimal estimation theory shows that speckle smoothing is best done before non-linear compression, and edge enhancement filtering is more effective after logarithmic compression. . For these reasons, the sequential post-detection image processing method in a conventional B-mode ultrasound imaging system cannot obtain the best results for both speckle smoothing and edge enhancement with opposing effects.
[0021]
The present invention provides adaptive B-mode image enhancement using post-detection image processing based on parallel signal paths. The parallel signal path according to this processing method includes a high-pass filter path that selectively enhances edges in the image and a low-pass filter path that selectively smoothes only the background speckles. Compared with the conventional sequential detection post-image processing method, the processing method according to the present invention mainly requires one additional logarithmic operation and can be implemented in hardware or software.
[0022]
In an adaptive B-mode image enhancement device according to a preferred aspect of the present invention, a low-pass filter for smoothing speckles and a high-pass filter for edge enhancement are arranged in parallel signal paths connected to the output of the envelope detector. The The signal in the signal path of the high pass filter is logarithmically compressed before high pass filtering. The signal in the signal path of the low pass filter is logarithmically compressed after low pass filtering. Respective weighting factors are added to the low-pass filtered signal and the high-pass filtered signal by the adaptive weighting means. The adaptive weighting means can take the form of a processing unit or a look-up table. The low pass filtered weighted signal and the high pass filtered weighted signal are summed and then optionally input to an anti-aliasing low pass filter prior to decimation and scan conversion.
[0023]
In general, the entire adaptive weighting and summing operation can be performed in the form of a look-up table if both low-pass filtered and high-pass filtered signals are available as inputs.
[0024]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 7 shows an adaptive B-mode image enhancement device 54 according to the present invention. An adaptive B-mode image enhancement device 54 is provided in the B-mode processing device to receive output from an envelope detector (ie, detector 46) of the type shown in FIG. In accordance with a preferred embodiment of the present invention, the adaptive B-mode image enhancement device 54 has a parallel path configuration that combines the advantages of the conventional configuration and eliminates the disadvantages of the conventional configuration. This adaptive B-mode image enhancement device 54 is not just an adaptive edge enhancement filter such as that shown in FIG. 6, but rather a more general post-detection image enhancement method including speckle reduction filtering and logarithmic compression. is there. Also, in the most general form shown in FIG. 7, the adaptive B-mode image enhancement device 54 is not limited to one-dimensional processing or vector-by-vector processing, and uses a two-dimensional filter for two-dimensional image processing. You can also
[0025]
According to the method of the present invention, the output from the envelope detector in the B-mode processing device is supplied to the first logarithmic compression means 48a and the first low-pass filter 52a, and the first of the adaptive weighting means 56 is used. Supplied to the input. Logarithmic compression function 48a is preferably provided in a software programmable look-up table. The logarithmically compressed data is supplied to the edge enhancement high pass filter 50 and also to the second input of the adaptive weighting means 56. The logarithmically compressed and high-pass filtered data is supplied to the multiplying means 58 a and to the third input of the adaptive weighting means 56.
[0026]
The low pass filter 52a is preferably a finite impulse response (FIR) filter. The low-pass filtered data is supplied to the second logarithmic compression means 48 b and to the fourth input of the adaptive weighting means 56. The logarithmic compression function 48b is also preferably provided in a software programmable look-up table. The logarithmically compressed and low-pass filtered data is supplied to the multiplying means 58 b and to the third input of the adaptive weighting means 56.
[0027]
The adaptive weighting means 56 may use any combination of the five data inputs shown in FIG. 7 depending on the application. The weight assignment can be specified by either a set of formulas or a lookup table. The adaptive weighting means 56 outputs the first weight W1 to the first multiplication means 58a and outputs the second weight W2 to the second multiplication means 58b. The sum of the weights W1 and W2 is equal to 1. The multiplier 58a outputs a signal representing the product of the edge-emphasized signal output from the high-pass filter 50 and the weight W1. The multiplication unit 58b outputs a signal representing the product of the speckle-reduced signal output from the logarithmic compression unit 48b and the weight W2. The outputs of these multipliers are then added by adding means 60. When both compressed low-pass filtered and compressed high-pass filtered signals are used as inputs, the entire adaptive weighting function 56, 58a and 58b and the addition operation 60 are implemented in the form of a look-up table. It is also possible to do. The added signal is then supplied to an optional anti-aliasing low-pass filter 52b. Alternatively, the summed signal can be supplied to the decimation means prior to scan conversion or directly to the scan converter.
[0028]
In the simplest case, the adaptive weighting means 56 can assign a weight based only on the output from the low pass filter 52a. If the output is large (edges), the weight on the output of the edge enhancement filter is greater and smoothing (shading) is minimized. If the amplitude is small (speckle), the weight on the smoothed and compressed output should be relatively large.
[0029]
As another example, to detect a cyst boundary in an abdominal scan, the adaptive weighting means monitors the difference between the output of the edge enhancement high pass filter and the logarithmically compressed low pass filter. If there is a large difference (boundary), assign a higher weight to the output of the edge enhancement high-pass filter.
[0030]
Unlike the prior art, the adaptive B-mode image enhancement device 54 shown in FIG. 7 has two separate low-pass filters 52a and 52b. One low-pass filter 52 a is a smoothing filter and forms part of the adaptive B-mode image enhancement device 54. The other low-pass filter 52b is an anti-aliasing filter and is optionally provided between the adaptive B-mode image enhancing device 54 and the decimation means or the scan converter. The low pass filters 52a and 52b are separated because they perform best at different locations in the signal processing system.
[0031]
A constant tap FIR filter (a boxcar for 1D vector processing) is a preferred low-pass filter for speckle reduction. This is because the filter provides an average value of the detected envelope signal that is shown to be the maximum likelihood estimate of the backscatter energy located below.
[0032]
The edge enhancement high-pass filter 50 is not necessarily an FIR filter. A high pass filter can be implemented using an IIR filter because its data input has a compressed dynamic range so that the transient output of the filter is very small.
[0033]
The adaptive weighting according to the broad concept of the present invention includes detector output, logarithm-compressed detector output, high-pass filter 50 output, low-pass filter 52a output and logarithmically compressed as shown in FIG. Or one or more functions of the low pass filter output.
[0034]
Compared to the prior art, the adaptive B-mode image enhancement device 54 mainly requires one additional logarithmic operation and does not significantly increase the computational burden and hardware costs.
Further, the adaptive B-mode image enhancement device 54 can be implemented in either hardware (eg, discrete components or ASIC) or software (eg, digital signal processor or Pentium Pro).
[0035]
The above preferred embodiments have been disclosed for purposes of illustration. Various modifications and variations will readily occur to those skilled in the field of ultrasound imaging systems. All such changes and modifications are intended to be included within the scope of the claims.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating major functional subsystems within a real-time ultrasound imaging system.
FIG. 2 is a block diagram illustrating details of the pulse generation and reception subsystem included in the system of FIG.
3 is a block diagram illustrating a receiver and a detector that form part of a beam former and a B-mode processor included in the system of FIG. 1, respectively.
FIG. 4 is a block diagram showing a conventional configuration of an image processing method after sequential detection.
FIG. 5 is a block diagram showing a conventional configuration of an image processing method after sequential detection.
FIG. 6 is a block diagram showing a conventional configuration of an image processing method after sequential detection.
FIG. 7 is a block diagram illustrating an adaptive B-mode image enhancement method according to a preferred embodiment of the present invention.
[Explanation of symbols]
2 Beamformer 4 B-mode processor 6 Scan converter / display controller 8 Kernel 10 Operator interface 12 Main controller 14 System control bus 16 Scan control sequencer 18 Scan control bus 20 Transducer 22 Color monitor 24 System timing generation 48a, 48b Logarithmic compression means 50 High-pass filters 52a, 52b Low-pass filter 54 Adaptive B-mode image enhancement device 56 Adaptive weighting means 58a, 58b Multiplication means 60 Addition means

Claims (9)

In a B-mode processing device for an ultrasound imaging system,
Envelope detector,
First logarithmic compression means having an input connected to receive the output of the envelope detector;
High-pass filter means for edge enhancement having an input connected to receive the output of the first logarithmic compression means;
First low pass filter means for speckle smoothing having an input connected to receive the output of the envelope detector;
A second logarithmic compression means having an input connected to receive the output of the first low pass filter means; an output of the envelope detector; an output of the first logarithmic compression means; the high pass filter A signal output from the high-pass filter means and having an input means connected to at least one of the output of the means, the output of the first low-pass filter means, and the output of the second logarithmic compression means Output means for representing the sum of the value obtained by applying the first weighting coefficient to the signal output from the first low-pass filter means and the value obtained by applying the second weighting coefficient to the signal output from the first low-pass filter means. B-mode processing apparatus, comprising: adaptive image enhancement means, wherein a weighting factor is determined as a function of a signal received from said at least one output.  The adaptive image enhancement means has input means connected to the at least one output, and adaptive weighting means having first and second outputs for outputting first and second weighting coefficients, respectively, the adaptive weighting A first input connected to the first output of the means and a second input connected to receive the output of the high pass filter means, and the signals received at these first and second inputs. A first multiplier for outputting a signal representative of a product; a first input connected to a second output of the adaptive weighting means; and a second connected to receive an output of the second logarithmic compression means. A second multiplying means having an input and outputting a signal representative of the product of the received signals at these first and second inputs, and connected to receive the outputs of the first and second multiplying means, respectively. And has a first and a second input, B-mode processing unit of the first and second of which according to claim 1, wherein an adding means for outputting a signal representing the sum of the signals received on the input.  2. The B-mode processing apparatus according to claim 1, wherein the adaptive image enhancing means has a lookup table having an address input connected to the at least one output.  2. The B-mode processing apparatus according to claim 1, further comprising second anti-aliasing low-pass filter means having an input connected to receive the output of said adaptive image enhancement means. In a post-detection image processing method in an ultrasound imaging system,
Logarithmically compressing a series of digital samples representing an envelope to create a series of logarithmically compressed digital samples;
High-pass filtering the series of logarithmically compressed digital samples to produce a series of highpass filtered logarithm-compressed digital samples representing an edge-enhanced image;
Low-pass filtering the series of digital samples representing the envelope to produce a series of low-pass filtered digital samples with reduced speckle;
Logarithmically compressing the series of low-pass filtered digital samples to produce a series of log-compressed low-pass filtered digital samples; and a first weight for the series of high-pass filtered logarithm-compressed digital samples Generating an image enhancement signal representative of the sum of the applied coefficients and the series of logarithm-compressed low-pass filtered digital samples applied with a second weighting factor. A post-detection image processing method.  The first and second weighting factors are the series of digital samples representing an envelope, the series of logarithm compressed digital samples, the series of high pass filtered logarithm compressed digital samples, the series of low pass filtering 6. The post-detection image processing method of claim 5, wherein the post-detection image processing method is determined as a function of at least one of the digital samples and the series of logarithmically compressed low-pass filtered digital samples.  The post-detection image processing method according to claim 5, further comprising a step of low-pass filtering the image enhancement signal. In ultrasound imaging systems,
A transducer array having a plurality of piezoelectric transducer elements;
A beam forming apparatus having a plurality of beam forming channels, switching means for coupling the piezoelectric transducer element and the beam forming channel;
An envelope detector coupled to receive a series of digital samples representing the received beam from the beamformer and outputting a series of digital samples representing the envelope; the series of digital samples representing the envelope; Means for logarithmically compressing samples to produce a series of logarithmically compressed digital samples;
Means for high pass filtering the series of logarithm compressed digital samples to produce a series of high pass filtered logarithm compressed digital samples representing an edge enhanced image;
Means for low-pass filtering the series of digital samples representing the envelope to produce a series of low-pass filtered digital samples with reduced speckle;
Means for logarithmically compressing the series of low-pass filtered digital samples to produce a series of log-compressed low-pass filtered digital samples;
Represents the sum of the series of high-pass filtered logarithm-compressed digital samples with a first weighting factor applied and the series of logarithm-compressed low-pass filtering digital samples with a second weighting factor applied An ultrasound imaging system comprising: means for creating an image enhancement signal; and means for displaying an image derived from the image enhancement signal.  The first and second weighting factors are the series of digital samples representing an envelope, the series of logarithm compressed digital samples, the series of high pass filtered logarithm compressed digital samples, the series of low pass filtering 9. The ultrasound imaging system of claim 8, wherein the ultrasound imaging system is determined as a function of at least one of the digital samples and the series of logarithmically compressed low-pass filtered digital samples.

Images